All Episodes

[Episode #33] – Fracking Follies


The US Energy Information Administration (EIA) regularly updates its estimates for how much oil and gas might be recovered in the future, and at what rate. With the application of new technology from year to year, those estimates generally keep going up. But it’s important to remember that they are just estimates — and the devil is always in the details.

Our guest in this episode is a career geoscientist who has diligently delved into those devilish details. In his new reports, he finds that EIA’s Annual Energy Outlook 2016 seems to significantly overstate how much oil and gas might be recovered using fracking technology, with estimates for shale gas and tight oil production that exceed the estimates for how much of those resources are even technically recoverable. In this extended and technically detailed interview, we discuss EIA’s most recent forecasts and try to understand what’s realistic for future US hydrocarbon production.

Guest: David Hughes is an earth scientist who has studied the energy resources of Canada for four decades, including 32 years with the Geological Survey of Canada as a scientist and research manager. He developed the National Coal Inventory to determine the availability and environmental constraints associated with Canada’s coal resources. As Team Leader for Unconventional Gas on the Canadian Gas Potential Committee, he coordinated the publication of a comprehensive assessment of Canada’s unconventional natural gas potential.

Over the past decade, Hughes has researched, published and lectured widely on global energy and sustainability issues in North America and internationally. Hughes is president of Global Sustainability Research, a consultancy dedicated to research on energy and sustainability issues. He is also a board member of Physicians, Scientists & Engineers for Healthy Energy (PSE Healthy Energy) and is a Fellow of Post Carbon Institute. Hughes contributed to Carbon Shift, an anthology edited by Thomas Homer-Dixon on the twin issues of peak energy and climate change, and his work has been featured in Nature, Canadian Business, Bloomberg, USA Today, as well as other popular press, radio, and television.

On the Web: David Hughes’ page at Post Carbon Institute

Recording date: December 3, 2016

Air date: December 28, 2016

Geek rating: 9

Chris Nelder: David Hughes is an Earth scientist who studied the energy resources of Canada for four decades, including 32 years with the geological survey of Canada, as a scientist and research manager. Over the course of his career he's led projects to assess Canada's coal and unconventional gas resources. And over the past decade, Hughes has researched, published and lectured widely on global energy and sustainability issues in North America and internationally, publishing numerous reports, production projections and debunkings of outlandishly optimistic official forecasts. With the support of the US based Post Carbon Institute, He published a series of papers which attempted to critically assess the true potential for unconventional oil and gas, including "Drill Baby Drill" in 2013 which examined the prospects for unconventional oil and gas in the United States, "Drilling California" in 2013 which critically examined the EIA's estimates of technically recoverable tight oil in the Monterey Shale after which the EIA wrote down its resource estimate for the Monterey by 96 percent, and "Drilling Deeper" in 2014 which challenged the U.S. Department of Energy's expectation of a long term domestic oil and natural gas abundance with an in-depth assessment of all drilling and production data from the major shale plays through mid 2014. I covered several of those reports in my previous career as an energy journalist and I'll link to all of the above in the show notes. Hughes has a pair of new reports for Post Carbon Institute titled "Tight Oil Reality Check 2016" and "Shale Gas Reality Check 2016". The reports compare EIA's fracking forecasts for tight oil and shale gas production from the annual energy outlook 2016 which will refer to by its acronym AEO 2016, to the forecast from AEO 2014 and 2015, as well as to Hughes' own forecasts from his aforementioned reports. Now I'm going to ignore EIA's occasional protestations that its estimates for future production are projections and not forecasts, even EIA occasionally calls them forecasts. And although this might seem like an awfully narrow subject for a podcast I assure you that once you finish this episode you'll know far more about fracking than you did before and maybe more than you ever wanted to know. Dave is an old friend an analyst whose work I regard very highly and it's a great pleasure to finally have him on the show. So let's bring him into the conversation now. Welcome Dave to the Energy Transition Show.

Dave Hughes: Good to be here Chris.

Chris Nelder: So let's dive right into your new shale gas report and then we can turn to the tight oil report after that. So in AEO 2016 EIA assumes that total U.S. gas production will actually begin to grow strongly starting next year, 2017, even though the rate of drilling new wells has fallen 37 percent from the 2014 peak, and it forecasts that tight oil and shale gas production will collectively grow 88 percent from 2014 levels to all time highs by 2040, while drilling rates actually remain below 2014 levels. So what's the justification that EIA gives for this increase in production without a corresponding increase in drilling. And what's your view of that.

Dave Hughes: Well the EIA doesn't give any justification for that. Typically when they publish an AEO it's released in April. In this case the final release came out in September. And they don't publish assumptions until months later, so still haven't published the assumptions that are behind the AEO 2016 yet. So what I did in my review was looked at the assumptions for the AEO 2015 report which were published last September. You know I suspect I was kind of looking under the covers that they're assuming there's going to be continued increase in technology, increase in the productivity of wells. But really you know if you look back at the actual technically recoverable resources in their 2015 assumptions it really doesn't make any sense, and on a play level, they're counting on recovering more gas and oil than actually exists in many of those places.

Chris Nelder: OK. Well if they're not publishing their assumptions that makes it pretty difficult to do a critical evaluation of what their forecast is. Are they planning to publish the assumptions that underlie the 2016 forecast?

Dave Hughes: Oh, they will eventually. It's just a question of when. You know, the EIA doesn't release the play level forecasts unless you ask for them.

Chris Nelder: Okay.

Dave Hughes: There's the nice lady in the EIA that has sent me them every year, when asked. In October she said the assumptions would be published in October. I just checked this morning and there's still not out yet.

Chris Nelder: OK.

Dave Hughes: Maybe they radically ramped up their assumptions in what's recoverable. I don't know.

Chris Nelder: Or maybe they see you know high prices being a driver here but as you point out EIA sees this remarkable production growth happening with only a modest increase in natural gas prices from around $2 and 80 cents per million BTU today to around $5 from 2025 all the way out through 2040. And you know first of all that seems like a pretty strange price scenario to me. I mean it makes sense that gas prices would get back to around 5 bucks because that's a price that can support a fair amount of shale gas production unlike the prices we have today. But then holding it steady around $5 for the next 15 years, when a glance at the history of gas prices show its never that stable seems highly unlikely to me.

Dave Hughes: Yeah it does to me too, especially when you look at the intrinsics of the plays themselves. You know good shale gas plays are not found everywhere. We have, for example, three quarters of the production in AEO 2016 it comes from just five plays. So, yeah, that price scenario doesn't make a lot of sense to me.

Chris Nelder: OK. So in the report you explain that the industry has dealt with the low prices we've had for the past two years in two main ways. One is high grading, which means focusing drilling on the most productive parts of the plays, the so-called sweet spots. And two by applying more aggressive technology, such as increasing the length of lateral wells, using far more fracking fluid during fracking operations and using many times more proppant, which is sand or other small particles that hold the fractures open, once the rock has been fracked. But you claim that while these practices have improved the production of some individual wells, because each well can now drain more of the reservoir, they also reduce the number of locations available to drill and simply exhaust a play more quickly at a lower cost without substantially increasing the ultimate recovery from the play. So what evidence is there to support this claim? Have we actually reached these limits in some plays where it doesn't make sense to drill a new lateral?

Dave Hughes: Well if you look at the play as a whole and you know there's several different reports out there, you know the story looks pretty good. If you look at the average productivity of wells drilled in each year over time the average has gone up in most plays. And that's just a function of those two things that you mentioned. You know early on in a shale play people are drilling everywhere. And as the sweet spots become delineated they focus their efforts on the most economic parts of the play and that's what's happening. Also the technology has been getting better. But if you look at a recent report by IHS, you know their thinking is that about two thirds of the increase in well productivity is due to high grading sweet spots and one third is due to technology getting better. And if you disaggregate the plays, so if you look at them by county, you know, for example, if you take a play like the Barnett for example which is really where shale gas got started and you look at a county like Tarrant which is really in the core of the sweet spot, you can see that the average well quality there has started to decline. Same thing in the sweet spots of the Bakken for example and the Eagle Ford. And so that's indicating that you know the technology is not producing any more increases in well productivity, in fact, wells are becoming so closely spaced together that they're beginning to interfere with one another. And so productivity is beginning to fall. I've seen that in the best county in the Marcellus, in the Bakken, Eagle Ford and the Barnett, you know it's just a fact of life. How those plays are going to evolve over time.

Chris Nelder: So what is the actual specific evidence that you've got some communication between these laterals that it doesn't make sense to drill another one in between them.

Dave Hughes: Basically the new wells have a lower quality, you know, lower productivity than the older wells. So that would indicate to me that there could be interference or it could simply be that they're drilling in lower quality parts of that particular county.

Chris Nelder: OK. So for those who aren't experts in shale gas we should probably explain a little bit here about the typical performance of shale gas wells and how quickly their production declines once they start producing. Can you give our listeners a brief overview of that.

Dave Hughes: Right. You know shale gas really got started with the combination of horizontal drilling and high volume hydraulic fracturing that evolved basically starting in the late 90s in the Barnett Shale of Texas and it's kept on evolving over time. But typically the wells decline quite quickly. You know for example in the Bakken first year declines are in the order of 65 to 70 percent, and 3 year declines are typically between 75 and 85 percent in most plays. So if you look at you know all wells drilled up to a certain point in time and you measure the field decline. Field decline is a combination of older wells that are declining more slowly and new wells are declining very quickly, the field decline in play like the Haynesville for example is about 45 percent in first year, which means you have to drill enough wells to replace 45 percent of production just to keep production flat. And that's pretty typical in most shale plays.

Chris Nelder: Okay. So 45 percent in the first year is fairly typical for shale gas well?

Dave Hughes: Well that's for a shale gas play.

Chris Nelder: Okay.

Dave Hughes: Individual wells decline more quickly in the first year.

Chris Nelder: That's right, you were saying like 85 percent?

Dave Hughes: Yes 75 to 85 percent over the first three years.

Chris Nelder: Okay.

Dave Hughes: But the bulk of that decline is in year one. No 65 70 percent in year one is not atypical.

Chris Nelder: Okay. So yeah this phenomenon of having to drill an increasing number of wells to compensate for the decline of old Wells has been compared to a treadmill or to the Red Queen from Alice in Wonderland we have to run faster all the time just to stay in place. But at the same time when the production from old wells falls below a certain level they'll be shut in, so there's also a limit to how many new wells you have to drill to maintain overall production. You don't always have this ever accumulating decline rate behind you. So do we have the data yet to know what the equilibrium level is between new well production and declining older wells and well retirements, maybe for older players like the Barnett and the Haynesville?

Dave Hughes: Yea, if you look at the Barnett which is where it really all got started, there's been just over 20,000 wells drilled and there's only about 16,000 producing at this point, so 4,000 of those wells have been shut in. You know the typical thinking is 25 or 30 year well life. However that's going to vary quite a bit. And it's looking, in the Barnett for example, that well life may be more like you know 10 to 15 years. Most of these plays aren't old enough so we've really seen know the full life cycle of the well play out.

Chris Nelder: Right.

Dave Hughes: Again it depends. You know the Barnett has declined quite a bit in terms of production from its peak. So the lower the the field production the fewer the number of wells that you need to drill per year. But for example, the Barnett it probably requires now about 400 wells per year to stay flat.

Chris Nelder: Okay.

Dave Hughes: If you look at a play like the Eagle Ford they were drilling 30 600 wells per year when it peaked, and drilling is now down to maybe half of that, but production is falling like a rock. You know production in the Eagle Ford is down 31 percent from peak.

Chris Nelder: So this concept of how many wells you have to drill to stay flat is really sort of a fiction anyway because we're not staying flat in these plays.

Dave Hughes: Exactly, and right now, because of the price of gas, and oil for that matter, the plays are being high graded. So we're really drilling the most economic parts as we speak. And what's left after that is lower productivity rock. So, you know, you may have to drill 2 wells in five years time to get the production that one well will give you today.

Chris Nelder: Right.

Dave Hughes: Which means that prices are going to have to go up.

Chris Nelder: I'm glad you mentioned that because I was thinking about this. You know given the decline of the existing wells doesn't EIA's remarkably stable price forecast, which again shows gas prices holding steady for 15 years straight, imply that there won't be any problem drilling new wells that are just as good and profitable as the old wells. Because if there were a problem, for example if you saturated the sweet spots and had to move off to the less profitable periphery of the plays, then you'd expect that prices would have to go higher over time to keep drilling enough more costly less productive wells in order to push the total production up as their forecast shows.

Dave Hughes: Exactly. And you know you can already see that if if you disaggregate the plays by county so you can see that well productivity is starting to fall. Which means that, you know it's getting to the law diminishing returns in those sweet spot counties. And you know there's still quite a few places to drill in some of those plays like, the Bakken and Eagle Ford. But you know the writings on the wall. You know barring some miraculous technology that will allow you know even more technological improvement but basically geology wins over technology every time it's just a fact of Mother Nature.

Chris Nelder: Well if that's the case that we can see in certain counties that well quality is falling then clearly one of the two things must be wrong here. Either they're flat gas price forecast for 15 years can't be right or if that is right, they can't have this increasing production that they're forecasting that they can't have those two things at the same time.

Dave Hughes: No. Exactly.

Chris Nelder: OK so throughout the report you actually show how EIA's forecasts for production in most of the shale gas plays increased really substantially from AEO 2015 to AEO 2016 but without any apparent justification for doing so, and you conclude the report with some very pointed questions. So let's talk about some of those. First you ask why the forecasts for production from these plays grew so much between AEO 2015 and AEO 2016 when the play fundamentals have hardly changed. What are some of those fundamentals that you're talking about there.

Dave Hughes: We talked about some of them. But basically the well productivity by by county has changed a little bit because of technology that, you know, as you said earlier, better technology means longer horizontal laterals, more profits, so each well is basically draining a larger part of the reservoir. You know that's why productivity is going up but then basically you reduce the number of locations that you have to drill. So even though you do get it more quickly and at a lower cost we don't really increase the ultimate recoverable resource very much. So the geological fundamentals are what I measured in my "Drilling Deeper" report in order to come up with with a forecast of future production. Basically that's the number of occasions in a play. So if you look at the aerial extent of a play and assume 4 or 6 or 8 wells per square mile you can determine the number of available locations and then you measure the well productivity, the average well productivity, by county and assume that you know the nature of the industry is to drill the best locations first and then proceed into the lower quality parts of the play. And you know when I did drilling deeper I actually made several forecasts based on different scenarios of drilling rates because future production is wholly a function of how fast you drill and well quality. I didn't try to forecast price. You know we've seen what price forecasts look like. So my forecasts are basically based on drilling rate and well quality.

Chris Nelder: But you didn't see the justification in AEO 2016 for these significantly increased forecasts from the EIA over the last couple of years forecasts that they put out.

Dave Hughes: No. No because they only published the overall aggregate forecast. That's what most people see. They don't actually see the play by play forecast which is what I looked at. You know they took apart their overall forecast by play. And if you do that you see some pretty crazily optimistic projections going on in my view.

Chris Nelder: OK so let's dig into that. So first I'm going to try to explain a couple of terms here just in case some listeners may not be familiar with them. So the first is technically recoverable resources which means an amount of oil and gas that is thought to exist which can be recovered with today's technology but without regard for price or profitability. And there might be an amount of proved technically recoverable resources meaning that at least the resources has been found by a drill bit. And then there might be a much larger estimate of unproved technically recoverable resources which is really usually just a mathematical probability of resources that haven't actually been found by a drill bit. And then there's proved reserves which means an estimate of how much oil and gas is known to exist and is thought to be producible at a profit at current prices. So it's only the prove reserves quality that we should really be able to count on producing in most cases. Now in your report you show how in AEO 2016 EIA sees production from the Marcellus, the largest of the natural gas plays, rising 48 percent above current levels, ultimately producing 47 percent more gas than the EIA's own estimate of unproved technically recoverable resources. And in the Haynesville play, AEO 2016 forecasts that production will more than double from current levels, with the Haynesville ultimately producing 28 percent more gas than the EIA's estimate of unproved resources. And in the Barnett, whose production peaked way back in 2011 as you said and has been declining ever since, they see production suddenly reversing course and that it's going to recover 145 percent more unproved resources than the EIA estimates even exist, and will end at a near all time high production level in 2040. I mean these are very strange things to forecast, to say the least. What is going on here?

Dave Hughes: I wish I knew. And you know if you look at that at the 2040 production forecast for the plays that you just mentioned, and that would imply that not only do you recover 145 percent more resource than they think exists in a play like the Barnett, but because production is so high when you exit 2040 they're implying that there's still you know vast amounts of recoverable resource left after that.

Chris Nelder: Right because production doesn't just fall off a cliff there is a tail.

Dave Hughes: Right. And you know my thinking just the way the geology was distributed for these places that you know you won't have a bell shaped curve of production it will have a long tail you know so steep steep ramp up to peak and then a long tail. So an asymmetric production profile. But having a radical ramp up at low prices you know when we've already seen some of the top counties exhibit declining well quality really doesn't fit with the fundamentals of my view.

Chris Nelder: I just don't understand how they could be forecasting production that is so much greater than their own estimate of unproved resources. I mean surely they're aware that there's a disconnect there. I mean do we have to assume that they're going to come out with some new assessment of larger unproved resources that will then sort of match up with this new production forecast, or, I mean I don't understand how they even get there.

Dave Hughes: Yeah. I suspect that that's what they will do. You know very few people actually get the spreadsheet of the play by play forecasts and without that you can't tell you know what's going on in these forecasts. So I make a point of doing that each year and that's the subject of these reports to kind of look at the credibility of this.

Chris Nelder: Right.

Dave Hughes: But the U.S. government I think is into you know good news stories. You know, energy independence Saudi America. We can build LNG export terminals and the price of gas is going to be cheap, you know for the foreseeable future, you know that's sort of the good news that makes you know everybody in Washington happy. But what is really going to happen when you dig into the geology is what I've been looking at.

Chris Nelder: OK. So I think the bottom line on this report is your final question how can overall shale gas production increase by 31 percent in AEO 2016 compared to AEO 2015 while assuming that gas prices are 20 percent lower over the 2015 to 2040 period. I mean is this, as you say, sort of an expression of a belief that future technology will produce more gas in the future at lower prices than it does today?

Dave Hughes: I suspect it is. You know I can't really think of any other explanation for it.

Chris Nelder: I mean they're not offering an explanation.

Dave Hughes: Well unless you know their assumptions for AEO 2016 are published sometime soon, maybe it will be some sort of an explanation in there. I'm really looking forward to it, but I couldn't wait. You know I wanted to get these reports out.

Chris Nelder: Well I mean yeah it's a 2016 report and 2016 is almost over. So yes I mean this is just a head shaker here. I can understand why you're taking the effort to look at this play by play because clearly something is amiss here. I mean throughout the report you not only compare the forecasts for each play from AEO 2016 to AEO 2015 and AEO 2014, but you also compare their forecast to your own more likely forecasts from your 2014 "Drilling Deeper" report, and you compare all those to the actual data that we now have for the two years since you wrote it. So on the whole how would you rate your own forecasts? Were they reasonably accurate and were they better or worse than EIA's?

Dave Hughes: Well, I'm obviously biased but I, you know, in actual fact I was a little bit too optimistic in 2014 for some of those plays.

Chris Nelder: And your forecast were lower than theirs.

Dave Hughes: And my forecast were lower than theirs. You know what I didn't project was the extreme drop in drilling rate and rig counts. So they dropped more than my most likely drilling rate, and therefore production was below my most likely forecast. So you know in general I was closer than the EIA but I was still a little optimistic on some of them.

Chris Nelder: I guess we have to assume that that was because neither you nor EIA expected this massive price drop that started in the middle of 2014 right around the time you're writing that report, like nobody knew that prices were going to fall as far as they did as rapidly as they did and then stay down for two years.

Dave Hughes: Yeah, that report was basically it was wrapped up almost exactly at the time that the world price of oil collapsed. So that wasn't foreseen.

Chris Nelder: Yeah. OK. So again, I guess you can be forgiven for being overly optimistic in your drilling forecast because your forecast was, as you said earlier, it was based on an expectation of some drilling rate. It was not based on a price forecast. EIA's forecast presumably were based, or at least within the context of their price forecast, but they didn't see prices crashing either.

Dave Hughes: No they didn't.

Chris Nelder: Okay. So let's turn to your companion report "Tight Oil Reality Check 2016". As with the shale gas report, it shows that EIA is forecast for tight oil production have continued to go up in recent years. But back in October 2014 when you wrote your "Drilling Deeper" report, you thought EIA's forecast was already too high, even back then, and you found that EIA's estimate of recovery from the Bakken and Eagle Ford plays in AEO 2014 was 42 percent higher than you're most likely drilling scenario. But now you say that EIA's estimate in AEO 2015 was 91 percent higher than yours and its estimate in AEO 2016 is 115 percent. So what's the deal? I mean are you just failing to keep up with the times here or what?

Dave Hughes: Well if you look at those two plays, the Bakken and Eagle Ford, actually the EIA did quite a reasonable job I thought in Eagle Ford. You know there was a paper that was published along with AEO 2014 that looked fairly similar to my methodology. They took it apart by county and looked at well quality by county, which is the way I do it. And actually their forecast with the Eagle Ford has been going down in successive years and now it's down almost to mine which I published in 2014. The Bakken on the other hand has just been widely increasing. This latest year AEO 2016, they've increased recovery by 2040 to a little over 18 billion barrels. That's close to triple. So the Bakken and maybe down, now I think it's down about 18 percent from its December 2014 peak but they forecast a massive resurrection to somewhere over double the 2014 peak. And you know again exiting 2040 at very high production levels. So i'm really scratching my head about it. I looked into the 2015 assumptions and they're assuming a huge resource from the Three Forks. You know a lot of the production is out of the Middle Bakken and some out of the Three Forks in the top counties, but there's a huge area of Three Forks that hasn't been drilled that they're assuming a vast resource from. So we know if you drilled over 100,000 wells and they turned out to be productive you know maybe that's what they're thinking. But the real problem in those two places the Bakken, you know what they did to that this year, not the Eagle Ford.

Chris Nelder: Right. So just to review the data here and in a little detail you note in your report that only two plays in the Permian Basin, the Spraberry and the Wolfcamp, are not below their production peaks already. All the other tight oil plays peaked and declined at various points in the past two years. The two largest plays, the Bakken and the Eagle Ford, which made up about half of all tight oil production around the middle of this year actually peaked in March of 2015. The Bakken is now 18 percent down from its peak as you said. And the Eagle Ford is down 31 percent from its peak but in AEO 2016 EIA forecasts are remarkably turn around for the Bakken rising just under one million barrels a day in 2018 to nearly 2.5 million barrels a day by 2033. But you see a gradual decline from 2015 onwards. So there's this big divergence now between your model and theirs. And for the Eagle Ford AEO 2016 forecasts a long shallow decline as you were saying, that long tail with production gradually falling from around a million barrels a day today to about three quarters of a million barrels a day by 2040, But you see a much more symmetrical bell curve of production for the Eagle Ford where it just falls to under half a million barrels a day at that point and has a total recovery of about 1.7 billion barrels less than EIA does. So what do you think accounts for the discrepancy between your model and theirs? I mean your methodology is very transparent. You look at the rate of drilling, the number of potential wells that can be drilled, how productive those wells will be and the rate at which those wells in that field decline. I guess I don't know if EIA actually is equally transparent about all those factors that go into its modeling. I mean what's the difference between your methodology and theirs to produce these kinds of divergent forecasts?

Dave Hughes: Well if you look at my forecast published in 2014 you'll see that I estimated the drilling rates would be quite a bit higher than they actually turned out to be. You know I overestimated the near-term production in Eagle Ford. So I assume drilling rates over 3,500 wells per year declining gradually into the future. But in fact drilling rates are down below eighteen hundred wells per year right now. So production fell quite a lot quicker than I expected but that saves drilling locations for later. So if I was going to redo that, you know, I would correct the near-term, and make it match existing production and increase the longer term forecast because you know you can drill locations now or you can drill them later and it looks like drilling them later is what's going to happen in the Eagle Ford. So actually the EIA's forecast, you know, was a little optimistic in my view, but it's reasonable projecting a higher production further out. And I do the same thing, You know if I was redoing the Eagle Ford in light of what's happened.

Chris Nelder: For the Eagle Ford, Yes, but do you think their forecast for the Bakken is realistic, going from a million barrels a day in 2018 to 2.5 by 2033.

Dave Hughes: I think is wildly optimistic. The Eagle Ford was not too bad, but they really made up for it in the Bakken. You know what would drilling rates have to be in order to double production from the peak rate. You know, peak rate, I think they were drilling somewhere around 1200 wells a year, maybe a little more than that. So off the top of my head. So to double, you know, not only are you drilling a lower quality rock, so you need more wells in the future for every well you're drilling today, to get that much production you probably have to be you know looking at four to five thousand wells per year in order to get there. And you'd run out of available locations way before you'd ever reach 20 40 production to go into terminal decline.

Chris Nelder: Well, not only that, I mean to reach that kind of drilling rate you'd have to just have such a mobilization of rigs and personnel and trucks to haul water and sand and fracking fluid and everything else. I mean you would almost certainly have, at least in North Dakota, an explosion of costs. And then you'd be off from your price forecast anyway.

Dave Hughes: Oh yeah. Yeah you'd definitely have to have the price to justify all the capital that would take. I mean we're looking at, you know, six to seven million a well. One of the reasons it's only six to seven million is because the service companies really lower their rates, you know when the price of oil fell.

Chris Nelder: So what was the per well cost like before the price crash in middle 2014?

Dave Hughes: They were higher.

Chris Nelder: OK.

Dave Hughes: You know they may be up around 8 million.

Chris Nelder: OK.

Dave Hughes: But we're also investing you know more into technology for these wells, so intrinsically they they're putting in more more water, more profit, drilling longer horizontal laterals. So if the Red Cross went back up to what it was two or three years ago we're probably talking you know nine million dollar wells. You know with the increased technological effort that's going into them.

Chris Nelder: So when EIA does come out with its assumptions underlying this 2016 forecast, do we think that they're going to answer these questions that when we go looking at the assumptions particularly for like a play level like the Bakken, we're going to find out just how many wells they think can be drilled and how productive the wells are and what the decline rate of the wells will be?

Dave Hughes: They'll certainly have an estimate of the number of wells. You know, the well density, the aerial extent, not necessarily decline rates.

Chris Nelder: Well if they don't have the decline rates then I guess they're not going to be publishing like a full type curve for each well, so what are they going to be reporting? Just like the initial production rate? And that's it? Or.

Dave Hughes: Well you know every year oil and gas assumption documents about 20 pages long. And there's a table of all of the different play areas, well density, numbers square miles of the play that you can calculate, how many locations are an average productivity. So that will probably be what we'll get. It's just a table, which is updated. You know I use that table from 2015 in my reports. It's going to be interesting to see what that table looks like when they get around to publish it.

Chris Nelder: Well you might have to write another report then to compare that side by side with the other one. So as your report shows, underlying EIA's production forecast is a price forecast that's pretty bullish. It sees oil going back to around $80 a barrel by 2020 and $100 a barrel before 2030. That's obviously considerably higher than the futures market is currently forecasting or that a lot of the typical data agencies, I don't think Wood Mackenzie or IHS or any of those other agencies are forecasting a price recovery like that. So do they explain how they arrived at that price forecast?

Dave Hughes: Well you know they have basically what they call a national energy modeling system. So they put a bunch of things like GDP growth forecasts into that, different parameters, and push a button, and they get a price forecast. It's a bit of a black box. So I'm not sure that we'll see any kind of details beyond that in terms of how they arrived at it.

Chris Nelder: So your report compares production estimates for the top nine major plays, plus an other category for all the rest. And you do that for AEO 2016 to AEO 2014 and to AEO 2015, and you find that the AEO 2016 estimates are pretty different both up and down from 2014 and 2015 forecasts. You have any idea why?

Dave Hughes: You know again, it all boils down to the assumptions. And in actual fact the bulk of them are up in AEO 2016, with the exception of the Eagle Ford which is down a little bit. But most of the plays are up. Hence you get that 31 percent increase over 12 months between 2015 and 2016.

Chris Nelder: But the EIA doesn't say why they think that things are going to be so much better in this report than they did in the last two years?

Dave Hughes: Not in detail, no. And the assumptions document will tell the tale, you know when that's released. Hopefully it will be released before AEO 2017 comes up.

Chris Nelder: Hopefully. So you also find that the AEO 2016 estimates for ultimate recovery are in some cases considerably higher than USGS' mean estimate of technically recoverable resources, especially for the Bakken, and you know it's never really been that clear to me if there is a relationship or what that relationship might be between EIA's technically recoverable resource estimates and the estimates from USGS. Do you know like what the connection might be between those or do they explain why their estimate for this is higher than USGS's?

Dave Hughes: Not that I've seen. You know, in general I would, you know, view the USGS work as being more credible than what the EIA does. And typically they're lower, like for the Marcellus that USGS came out an 84 TCF number. Now the EIA's, I don't have it exactly offhand what their estimate for the Marcellus is but it's over 200 TCF. It's probably more like triple the USGS estimate.

Chris Nelder: But is there any like I don't know formal methodological relationship between EIA's estimates for TR and USGS' or or not they're totally separate?

Dave Hughes: I think that they are totally separate. I don't know the EIA does it.

Chris Nelder: I mean these are both U.S. government agencies. This seems like a very strange way to do things.

Dave Hughes: Yeah it does. When you consider the importance of of getting it right.

Chris Nelder: Well yeah. So I mean this has puzzled me for years, for literally many years. And I think I've even looked to try to find any explanation published by either agency as to why their estimates don't line up or why there is no apparent relationship between them. I mean, do you have any insight on that at all?

Dave Hughes: Well the the people that do the USGS estimates are a Ph.D. level statisticians and geologists for the most part. So I, you know, I would put my money on them in terms of having more rigorous methodology. When it comes to the EIA i'm not exactly sure how to come up with those numbers.

Chris Nelder: OK.

Dave Hughes: I'm just looking at the Marcellus now, and their 2016 report they assume we're covering 215 TCF from 2014 to 2040. And the USGS' mean estimate was 84 TCF of undiscovered technically recoverable.

Chris Nelder: So. So. So so EIA is basically forecasting about two and a half times as much gas is going to be recovered from the Marcellus as USGS thinks exist in undiscovered technically recoverable resources.

Dave Hughes: Right. And, in fact, the EIA's estimate of unproved technically recoverable resources.

Chris Nelder: Wow.

Dave Hughes: Plus the fact that the Marcellus is exiting 2040 at a near all time record levels, you know, which indicates that there's got to be, you know, another hundred or two hundred TCF of recoverable gas there as well.

Chris Nelder: Wow. I mean this is just so unsatisfying, you know. If you actually care about the numbers. I'm not saying that EIA's forecasts here are wrong, because I don't know. I don't have any ability to know. I am saying that it's very unsatisfying the way they've gone about justifying or explaining their assumptions here. I mean this is they're saying things that clearly don't match with what they have said in the past and what other agencies have said. And they're not explaining the divergence.

Dave Hughes: And what the fundamentals of the well production data say.

Chris Nelder: Exactly.

Dave Hughes: If you care to go to that depth.

Chris Nelder: Yeah.

Dave Hughes: But the worrying thing is industry is making policy based on this. We're building LNG export facilities, assuming cheap abundant gas for the foreseeable future. And if those forecasts aren't realistic, all that capital investment is is just stranded. And it puts longer term energy security at risk. You know all the decisions that are being made now based on those assumptions.

Chris Nelder: That's a great point. And I mean I have to assume that the companies that are stepping up to spend a couple billion dollars on a new LNG export facility have to be reasonably confident about the assumptions on which they're basing that investment. I certainly hope they've got more to go on than this EIA forecasting.

Dave Hughes: It's actually more like 10 billion to build a big LNG export plant, upfront. And the way some of those projects are working you know it's basically cost plus. So all they're doing is providing a service, you know whatever the domestic price of gas is they add X cost to that. So they're just providing a service to liquefy the gas and ship it. And you know whatever the input prices is is the responsibility of the people that signed the contracts to export that LNG. So there's just not that much risk in price spikes to the actual LNG facility owners. You know that risk is so on the the people that sign long term take or pay contracts.

Chris Nelder: Well good luck to them I guess. So I think the bottom line on both of these reports is that you think there are problems with EIA's national energy modeling system known as NEMS. It's acronym. So why don't you just go ahead and tell us what you think the issues are there.

Dave Hughes: Well you know I tried to go through as objectively as I can to look at what might a reasonable outlook be, just to kind of factor in you know a reality check which is what the reports are called, that all may not be quite as sweet as we think and that should be a wake up call, you know to utilities that are converting to gas big time. And the likely future prices of electricity, if price spikes, which is quite likely to happen. And if you look at what happened in the northeast a couple of years ago in January, the price of gas spiked to $100 a million BTUs from about $2 is what it is selling for today in the northeast. So historically natural gas has been a pretty volatile commodity. As recently as mid 2008 it was selling for over 12 dollars. So to assume that it's going to be $5 out to 2040 is a pretty optimistic assumption in my view.

Chris Nelder: Well clearly. And with no volatility in their price forecast.

Dave Hughes: Right. Making price forecast is a bit of a mugs game. And volatility is a pretty hard thing to forecast. So if you're going to get the gas out of the ground at the levels that they're projecting, prices are going to have to go a lot higher than $5. And because of the fact that I mentioned right at the beginning that 5 shale gas plays produced three quarters of the gas that the EIA is planning on recovering by 2040, you know we're basically drilling off the best parts of those plays as we speak. So that tells you that longer term prices are going to have to go up and supply likely go down even with considerably higher prices. So that should all be factored into the energy decisions that we're making today.

Chris Nelder: Well and this is really the point where we have to point out that most of the additional demand for gas, in the US anyway for the last couple of years, has been driven by the power sector. That gas is being used to generate electricity. And if the price of gas rises substantially from where it is now, gas fired power generation is going to be out of the money relative to renewables certainly because right now renewables are already knocking on the door of gas fired power as, or close to price parity, and it will probably become more expensive than coal depending on how high it goes. I mean I think coal fired power now in the U.S. goes for around 6- 7 cents per kilowatt hour gas fired power is closer to 5 cents a kilowatt hour. If we double the price of gas or even more, as I guess it would have to be given this forecast that you're looking at if gas prices go go up to let's say $8 per million BTU and gas fired power just kind of making numbers off the top of my head here, but let's say gas fired power goes back to eight cents per kilowatt hour, you would have to expect that we're going to start shutting down these gas fire generators. We might go back to coal and renewables are going to just absolutely run away with all new capacity. Like we're not going to be building any new gas fired power plants at that point. It's all going to be renewables. So that alone undercuts this notion of having a higher gas price in the future that would justify this production.

Dave Hughes: Yeah well it's even a little worse than that, because we're actually shutting down coal plants and dismantling them. So they're gone. So we're creating an inelastic demand for gas. I mean if you want power you've got to use the gas. There's no longer an alternative with coal. You know there is now, but that's gradually going away. You know right now gas is more competitive than coal. But you know a lot of that coal capacity is is coming off line. So there's really no choice unless you have an alternative such as renewables. So you're kind of stuck between a rock and a hard place if you get a big gas price spike and you have no alternatives to it.

Chris Nelder: Yeah I mean you just have to add a spike grid power prices and that's it.

Dave Hughes: Yeah. Which is exactly what happened in New England.

Chris Nelder: Right now we should note that sense that polar vortex freeze that you're talking about there in New England, during which by the way several coal plants couldn't function because the piles of coal actually froze solid and they couldn't literally couldn't get it into the furnace. Since then we've built quite a bit of new pipeline capacity in the Northeast, and some new storage capacity. So I think the hope at least within the industry and the utility sector in the Northeast is that they've got more of a buffer now and more of a fallback position than they did in the polar vortex. But of course it's not unlimited.

Dave Hughes: Yeah you're exactly right. There has been a lot of pipelines capacity built. And you know the Utica is a play that has really evolved quite rapidly in the last couple of years, feeding into the same market as the Marcellus. But people have ambitions. Canadians in Nova Scotia are looking at building an LNG export facility for some of that northeast U.S. gas. So you know people are adding demands for that as well as we speak.

Chris Nelder: Yeah this whole LNG export thing really deserves a whole nother discussion because I mean these are as you say very expensive plants. These export facilities, these liquefaction facilities and export terminals are very very expensive. Billions of dollars. So it takes a long time to build them. And they were, all the ones that we're building now were on the drawing board back when you could expect to fetch 10 or 12 dollars per million BTU for the gas when it landed in let's say Asia. That's no longer the case. Gas prices have fallen so hard around the world that as far as I know you can't actually ship a cargo of gas from the U.S. to Asia at a profit right now.

Dave Hughes: No you can't. Absolutely. You know when everybody was really keen on LNG back in 2013 the price of LNG landed in Japan was $18 I just checked it. But now it's below six.

Chris Nelder: Wow.

Dave Hughes: And in essence to get LNG from say New Orleans to Asia is somewhere between five and six U.S., just for the liquefaction and shipping costs. So if it costs you four at the terminal on the Gulf Coast and it cost of six to ship it you need 10 just to break even at selling for the less than six.

Chris Nelder: Right.

Dave Hughes: So it doesn't make a lot of sense right now.

Chris Nelder: Well I got to wonder what they're thinking up there with that export terminal in Nova Scotia then.

Dave Hughes: Yeah me too. You know we'll see where that goes. I don't think it's going to go very far. If you look around the world at what's happened with gas production, in fact gas production has gone up more rapidly in the rest of the world than it has in North America. So I you know I think that the low prices in Asia and Europe are probably going to be here for a while.

Chris Nelder: And I think most of that new gas production that you're talking about has actually been Australia, Indonesia and Russia. Right.

Dave Hughes: A lot of new LNG out of Australia. For sure some out of Russia. But, you know, overland pipelines from Russia to China. You know if you look at the thermodynamics of LNG, it takes about 20 percent of the gas just to provide power for the liquefaction and shipping and re-gasification part of the process. So you know from an energy conservation point of view LNG is an expensive way to move gas, both in terms of money and in terms of energy.

Chris Nelder: A 20 percent parasitic load right off the top. Basically.

Dave Hughes: Yeah. Right off the top.

Chris Nelder: Interesting. So I suppose we should take just a minute here before we wrap up to talk about the Wolfcamp shale in the Permian Basin since that's been in the news lately. I explained the basics in the News segment to episode 31. It was basically a reassessment by the U.S. Geological Survey of the Wolfcamp shale which suggested that it could hold as much as 20 billion barrels of oil, 16 trillion cubic feet of natural gas and 1.6 billion barrels of natural gas liquids. But contrary to many press reports, it wasn't actually a discovery but rather a reassessment of an old play that we've been drilling for a long time that has some 7,000 wells currently producing on it. And the USGS estimate was really just for a 50 percent probability of undiscovered technically recoverable resources that could be recovered from the play if it were redeveloped using horizontal drilling and fracking. But we don't actually know, yet, what price it would take to make these undiscovered resources into commercially viable proved reserves and we might not know that for years. So is there anything you want to add on the Wolfcamp story?

Dave Hughes: Where you're right. You know ahead to another look at that just a couple of days ago. And there's been over 13,000 wells drilled since the 50's in the Wolfcamp. It's produced a billion barrels of oil since the 50's and about 5 trillion cubic feet of gas. So you know we've known about it for a long time, so it's not a not a new discovery as the press made it out to be. You know as you say it's a very thick formation with several benches in it that look like they can be developed with horizontal drilling and fracking. But what would it cost? That's the key. And the USGS provides an estimate of the ultimate recovery of wells, you know by bench in the Wolfcamp for their resource assessment. So actually I took Art Berman's numbers, you know 7 million per well, his operating cost of about $12 a barrel. And I looked at the EURs from the USGS study.

Chris Nelder: The EURs, just for people who don't know, it's the the estimated ultimate recovery.

Dave Hughes: Per well. And just, you know, correct some costs. It turns out you need to drill 169 thousand wells at seven million each to produce 20 billion barrels of oil with a $12 per barrel operating cost. So if you assume the price of oil is $45 and there's 20 billion barrels that might be recoverable that's nine hundred billion. But just to get it out of the ground to drill those hundred sixty nine thousand wells you need 1.4 trillion. So you know just on a straight up capital investment costs you lose 500 billion and that doesn't count the oil fees and any taxes you have to pay. So you know I wouldn't count on getting that oil out anytime soon. Although certainly, you know the Wolfcamp, it is producing oil and it has gone up quite a bit in terms of production. And it will continue to produce oil. But I certainly wouldn't hold my breath on the 20 billion. Considering you know what is going to take in terms of wells, capital costs and so forth.

Chris Nelder: And again, I mean it just makes no sense to think about these kinds of numbers, you know recovering $20 billion barrels from this play over how many decades into the future without also considering what's the likely demand going to be, and that has to be based on the number of vehicles that are still burning petroleum. And that has to take into account what's your forecast for electric vehicle adoption. Likewise it doesn't make any sense to talk about these kinds of shale production forecasts that you've been looking at here play by play without thinking about what's the demand likely to be from the power sector. And you can't think about that without understanding the growth of renewables and the price of renewables relative to fossil fuel generation and so on. And yet, none of this kind of integrative modeling is actually being done. Not that I'm aware of. EIA isn't doing it. Nobody's doing it.

Dave Hughes: No that's right. You know and that's exactly why I'm interested in trying to put together an objective evaluation of these kinds of forecasts because they're so important. You know people there are making huge investments based on them. So you know the closer they are to reality the less disruption down the road is going to be in terms of sustainable energy future.

Well Dave I really appreciate you taking the time to explain some of this for us and actually you know just doing the work in the first place because as you say it's extremely rare to find anybody who's looking at this stuff at a play by play level and really explaining it, but I've got to say it's a little scary to see the cover pulled back on this and to see how just wacky some of the forecasts are underneath the cover here. This is unsettling.

Dave Hughes: Yeah. You know if you look at who has the resources and who has the loudest megaphone it's really industry investor presentations, you know the next quarter. Government should be a little more objective but they seemed to be pretty optimistic.

Chris Nelder: Well yeah.

Dave Hughes: Politicians that are making the policy decisions, you know, don't really have a grounding in how believable some of this information actually is.

Chris Nelder: Well especially now that we have or we're going to have President Trump and he's apparently been talking to people like Harold Hamm from the CEO of Continental Resources and Rex Tillerson from Exxon Mobile to play key roles in his administration, which adds even less confidence to the realism of our plans going forward.

Dave Hughes: Yeah absolutely.

Chris Nelder: Scary stuff man. This is trippy. Maybe one when EIA does come out with the assumptions underneath it's AEO 2016 you can write a little report, once again sort of explaining it for the lay person and maybe we can get you back on the show at that point to talk about it.

Dave Hughes: Yeah. I'm waiting with bated breath to see what those assumptions are Chris.

Dave Hughes: Well thanks a lot for coming on the show Dave, I appreciate it.

Dave Hughes: You're very welcome.